336 1
(16) (17) (18) (19)
the previous acidic conditions (37%).8 Reduction and other reactions of norbornen-7-one proceed similarly, although the reported stereospecificity is somewhat lower: R. S. Bly, C. M. Dubose, Jr., and G. B. Konizer, J. Org. Chem., 33, 2188 (1968), and references cited therein. From examination of models it would appear that the axial hydrogens at C(2) and C(4) provide effective additional hindrance to approach to the carbonyl group of 15 from that side of the molecule, and this may account for the greater stereoselectivity observed in the bicyclooctenone system. H. Krieger, Suom. Kemistil. 8,38, 260 (1965). R. A. Appleton. J. C. Fairlee, R. McCrindle, and W. Parker, J. Chem. SOC. C, 1716 (1968). Assignment of this upfield signal to the anti proton at C(8) in 18 is required by the observed spectra of 16 and 17 taken with the fact that 12b is known' to react under solvolysis conditions with retention of configuration. It is also in line with earlier assignments for the anti proton at C(7) in various norbornene epoxides: K. Tori, K.Kitahonoki, Y. Takano, H. Tanida, and T. Tsuji, Tetrahedron Left., 559 (1964); J. E. Franz, C. Osuch, and M. W. Dietrich, J. Org. Chem., 29, 2922 (1964); R. Huisgen, L. Moebius, G. Mueller, H. Stangl, G. Szeimies, and J. M. Vernon, Chem. Ber., 98,3992 (1965). The only previous observations concerning bicyclo[3.2.l]oct-6-ene epoxides known to us are those of P. R. Jefferies, R. S. Rosich, and D.E. White [Tetrahedron Left., 1853 (1963)] on oxiranes derived from the diterpene alcohol beyerol and its esters. This was the original report that epoxides cause an upfield shift of suitably located nearby hydrogens, but in this work the unique upfield signal was attributed to the C(8) proton syn rather than anti to the epoxide ring. The subsequent results with norbornene epoxides cited above, along with our present findings with 16-18, suggest that this original assignment by Jefferies et al. should be reversed. in contrast, reductionbf 12b with lithium aluminum deuteride in refluxing ether gave -15% inversion and -85% retention from similar analysis of the derived epoxide. H. C. Brown, "Organic Syntheses via Boranes", Wiley, New York, N.Y., 1975. A small amount (8%) of aldehyde 21 is also formed, but this is believed to arise from inverted conformer 2e. See ref 3 for discussion. Such substitution also has an important effect in related ground state reactions: W. C. Agosta and s. Wolff, J. Org. Chem., 40, 1027 (1975).
(20) R. H. Shapiro, Org. React., 23, 405 (1976). (21) H.C. Brown and G. Zweifel, J, Am. Chem. SOC.,83, 1241 (1961). (22) S. Wolff, W. L. Schreiber, A. B. Smith, 111, and W. C. Agosta, J. Am. Chem. Soc., 94, 7797 (1972). (23) H. 0. House, "Modern Synthetic Reactions", 2nd 4,W. A. Benjamin, Menlo Park, Calif.. 1972, pp 412-414, and references cited therein. (24) Spectral assignments for these ketones are given by W. C. Agosta and S. Wolff, J. Org. Chem., 40, 1665 (1975). (25) D. Y. Curtin and D. B. Keilom, J. Am. Chem. Soc., 75, 601 1 (1953). (26) P. J. Boddy and E. W. R. Steacie,Can.J,Chem.,38,4576(1976); J.H.Current and B. S.Rabinovitch, J. Chem. Phys., 38, 783 (1963). (27) W. B. Hammond and T. S. Yeung, fetrahedron Lett., 1173 (1975), and references cited therein (28) Only a small steric effect is anticipated since for cis- and transhydrindan is only -0.3 k c a l h o l and actually favors the trans isomer. For discussion and references see E. L. Eliel, N. L. Allinger, S. J. Angyal, and G. A. Morrison, "Conformational Analysis", Wiley, New York. N.Y., 1965, pp 228-230. (29) A. L. J. Beckwith and G.Phillipou, Aust. J. Chem., 29, 123 (1976); A. L. J. Beckwith, Chem. Soc., Spec. Pub/., No. 24, 239 (1970); J. W. Wilt in "Free Radicals", Vol. 1, J. K. Kochi, Ed., Wiley, New York, N.Y., 1973, Chapter 8. These articles contain extensive references to earlier literature. (30) P. Livant and R. G. Lawler, J. Am. Chem. SOC.,98, 6044 (1976). (31) E. L. Eliel, N. L. Allinger, S. J. Angyal, and G. A. Morrison, "Conformational Analysis", Wiley, New York, N.Y., 1965, p 9, and references cited therein. (32) C. H. Bushweller and W. G. Anderson, Tetrahedron Lett., 181 1 (1972). (33) N. Ya. Buben, Yu. N. Molin, A. I. Pristupa, and V. N. Shamshev, Dokl. Akad. Nauk SSSR, 152,352 (1963); S. Ogawa and R. W. Fessenden, J. Chem. Phys., 41, 994 (1964). (34) K . U.lnaoid in "Free Radicals", Vol. 1 , J. K. Kochi. Ed.. Wiiey, New York. N.Y., 1573, Chapter 2 (35) See, for example, P. D. Hobbs and P. D.Magnus, J. Am. Chem. Soc.. 9 8 , 4594 (1976). (36) For a recent study of substituent effects that are interpreted as introducing barriers to conformational change in biradicals arising in a rather different kind of photochemical reaction see J. R. Scheffer, B. M. Jennings, and J. P. Louwerens, J. Am. Chem. Soc.. 98 (1976).
The Photochemistry of Chloroaromatic Compounds. Is "slr-Chlorobenzene" an Intermediate? D. R. Arnold* and P. C . Wong Contribution from the Photochemistry Unit, Department of Cheniistry, University of Western Ontario, London, Ontario, Canada N 6 A 5B7. Rereiced October 18. I976
Abstract: In 1973 Lemal and co-workers2dreported that irradiation of dilute solutions of chlorobenzene i n cqclohexane gave significant amounts of chlorocyclohexane ( 5 3 % ) . To account for this unexpected observation they proposed a mechanism involving hydrogen abstraction from cyclohexane by a phenyl radical which is complexed with a chlorine atom ("a-chlorobenzene"). The cyclohexyl radical and the chlorine atom then couple in the solvent cage. We have reexamined this reaction and o u r results lead us to propose a n alternative explanation. We present evidence that the chlorocyclohexane derives from the photosensitized addition of hydrogen chloride to cyclohexene and that the cyclohexene is formed by disproportionation of cyclohexyl radicals.
In 1973 Lemal and co-workers reported that irradiation of dilute solutions of chlorobenzene (I) in cyclohexane (11) gave chlorocyclohexane (III,53%) as a major product.2 They recognized that formation of this product could not simply be explained as the reaction of the phenyl and chlorine radical, which would result upon bond homolysis, since both the phenyl radical and the chlorine atom are relatively reactive radicals and would abstract hydrogen, predominantly from the solvent. Benzene was a major product but there was less than a n equivalent amount of hydrogen chloride. To account for this unexpected result they proposed a mechanism, the key step of which was a hydrogen abstraction from cyclohexane by the phenyl radical moiety of a chlorine atom-phenyl radical complex ( V , "r-chlorobenzene") giving a chlorine atom (complexed with benzene)-cyclohexyl radical pair in a solvent cage (Scheme I ) .
We were attracted to this area by the unusual nature of this suggested mechanism and particularly by the possibility that reactions involving this type of intermediate ( V ) might offer a synthetically useful procedure for functionalizing an unactivated, saturated, proximate alkyl position. For example, o-chloropropylbenzene ( V I ) would be expected to yield 3chloropropylbenzene (VII) since the hydrogen abstraction from the terminal methyl group by the phenyl radical moiety is favored in a six-membered transition state as shown in Scheme 11. This type of intramolecular hydrogen abstraction has been observed by Beckwith and co-workers using electron spin resonance (ESR) technique^.^ We recognized that this particular example puts a heavy demand upon the intramolecular reactivity of the intermediate VI11 since the hydrogen being abstracted is attached to a terminal carbon (i.e., primary hy-
Arnold, Wong
/ Photochemistry of Chloroaromatic Compounds
3362 Scheme I
sOll
'
1
70.
1
I l
L
J
E
m
Scheme I1 r
L
J
m
SZI
Figure 1. Product distribution of irradiated solution of o-chloropropylbenzene in cyclohexane vs. irradiation time.
Scheme 111 r
i
r
i
drogen), if the reaction were carried out in cyclohexane. Therefore, we use neopentane (IX) as solvent. The irradiation of o-chloropropylbenzene (VI) in neopentane (IX) solution led to a complex mixture of products; the major products were n-propylbenzene (X) and hydrogen chloride. Most significant for the present discussion was the absence of a detectable amount of 3-chloropropylbenzene (VII), nor were the other side chain chlorinated propylbenzenes (i.e., 1-chloropropylbenzene (XI) or 2-chloropropylbenzene (XII)) significant products. Furthermore, neopentyl chloride (XI I I) was only a minor (
